Abstract
In this study, we analyzed the effect of dispersant characteristics on the selective catalytic reduction (SCR) catalyst properties and de-NOx efficiency. For this, we measured the zeta potential and pH value of each dispersant, and compared the thermal properties of the dispersant through TG-DTA analysis. Also, the Py-GC/MS analysis results and the MSDS contents of the product were used to compare the components and molecular weight types of the dispersant. As a result, the higher the zeta potential, pH, and molecular weight of the dispersant, the more improved the dispersibility of the TiO2 slurry. Characteristics such as the rheology, sedimentation, and pH change, were studied to compare the dispersibility of the catalyst slurries, and the dispersion characteristics of the TiO2 slurries were confirmed by TEM. The SCR catalysts prepared varied based on the dispersant added, with the varying factor being the de-NOx efficiency between (250 to 450) °C depending on the dispersibility. The dispersant with the excellent dispersibility gave the highest efficiency of 84% or more at 250°C and 300°C, and the highest de-NOx efficiency of more than 92% at 350°C and 400°C.
Introduction
Nitrogen oxide (NOx) is one of the air pollutants generated during a combustion processes. NOx in the atmosphere is a causative agent of acid rain and photochemical smog; also, recent years, it has been known to act as a precursor of secondary particles. The selective catalytic reduction (SCR) method is known as an optimal method for removing NOx, because it has high de-NOx efficiency, it is economical, and has high stability.
1
The basic principle of SCR is that NOx is reduced to N2 and H2O by reacting it with NH3, and the reaction formula of NH3-SCR is as follows
2
:
Catalyst components (supports, active species, etc.) directly affect catalyst performance. In actual SCR catalyst manufacturing, in addition to these catalyst components, additives such as binders (both organic and inorganic), and dispersants are mixed with catalyst materials to maintain the actual catalyst shape and improve durability. In particular, the dispersant controls the agglomeration of solid substances in the catalyst slurry. Lee et al. reported that the strength and de-NOx efficiency of the catalyst were changed when a dispersant was added during the preparation of the SCR catalyst. 10 Wang et al. studied the effect of tungsten dispersion on the surface of titania catalyst, and reported that incomplete dispersion of tungsten oxide reduces the surface Bronsted acid sites and reduces the catalytic activity. 11 Chen et al. studied the low-temperature SCR activity of Mn/Ce/TiW catalysts and explained that the improvement of dispersion degree is the main reason for enhancing acidity, reducibility and SCR activity. 12 As the dispersion of catalyst materials is reported as an important factor in catalytic activity, studies to improve the dispersion of the catalyst material on the catalyst surface have been conducted. According to Ye et al., when the dispersibility of the active material is low, the activity decreases due to aggregation and crystallization as the catalyst active species increases. 13 In their study, they used N-doped TiO2 to improve dispersion, and reported an increase in both the dispersion and catalytic activity of the catalyst material. Zhang et al. improved the dispersibility of active species Cu and Ce by ultrasonic-assisted impregnation method during the preparation of CuCe/TiO2-ZrO2 catalyst, and reported that NO conversion was significantly increased. 14
In a catalytic reaction, the dispersibility of the catalyst component is an important factor, therefore, in other catalytic fields, various studies have been conducted to analyze the effect of the dispersant's properties on the catalytic activity.15,16 In the field of SCR catalysts, various studies have been conducted to improve dispersibility, however, few studies have been conducted to analyze the effect of physicochemical properties of dispersants on catalytic activity. Therefore, in this study, we investigated the effects of the physicochemical properties of the dispersants added during the preparation of the catalyst and nanoparticle TiO2 slurry on the de-NOx efficiency of the catalyst.
Experimental details
Characterization of the dispersant and TiO2 slurry
4 types of dispersants were used; BYK110, BYK180 (BYK, Germany), 368WB and 369SB (DYNEChemtech Corp., Korea). Each zeta potential was analyzed using a zeta potential analyzer (ELSZ-200ZS, Otsuka). The pH of each dispersant was measured by pH meter (inoLab pH7310, WTW). Thermogravimetric analysis were performed to check the thermal stability of the dispersants by heating 3 mg of dispersant from ambient temperature to 800°C at the rate of 20°C/min under nitrogen atmosphere (100 mL/min) using a TG analyzer (TA 55, TA instrument). Pyrolysis-gas chromatography/mass spectrometry (Py-GC/MS) analysis was carried out using a pyrolyzer (EGA/Py-3030D, Frontier-Lab.) directly coupled with a GC/MS (7890A/5975 Inert, Agilent). Table 1 details the condition of the Py-GC/MS analysis.
Py-GC/MS analysis conditions applied in this study.
To evaluate the dispersion characteristics of the dispersant added to the TiO2 slurry, we prepared the sample by mixing TiO2 10.0 wt.%, colloidal silica 89.0 wt.%, and dispersant 1.0 wt.%. For the sedimentation study, we measured the change in the suspension height of TiO2 slurry over time, and evaluated the sample that took a long time for the settling surface to reach an arbitrary height as having excellent dispersibility. Using Transmission Electron Microscopy (TEM, CM200, Pilips), we compared the dispersion shape of nanoparticles based on the type of dispersant added to the TiO2 slurry.
Preparation of catalyst slurry and SCR catalyst
Colloidal silica, a binder, and a dispersant were mixed to prepare a catalyst slurry to investigate the characteristics of the catalyst slurry and the de-NOx efficiency according to the dispersant. TiO2 44.8 wt.%, colloidal silica 49.8 wt.%, binder 1.0 wt.%, and dispersant 1.0 wt.% were mixed to form a catalyst, and the remaining contents were catalytically active material such as V2O5 and WO3. The prepared catalyst slurry was thermally calcined at 400°C for 1 h, and then pulverized to be used as an SCR catalyst to evaluate the de-NOx efficiency. Table 2 shows the specifications of TiO2 powder used in this study.
Specification of TiO2 nano powder.
Characterization of the catalyst slurry and SCR catalyst
To evaluate the dispersibility of the catalyst slurry, we used rheometry (ARES-G2, TA Instruments) to measure the variation of viscosity with shear rate, and also measured the pH. In addition, the slurry was calcined at 400°C for 1 h, then pulverized, and the specific surface area and pore characteristics of each catalyst were confirmed through the N2 adsorption–desorption method (3Flex version 3.02, Micromeritics)
SCR catalyst de-NOx efficiency evaluation
To investigate the effect of the properties of the dispersant on the de-NOx efficiency of the SCR catalyst, we evaluated the de-NOx efficiency using the SCR catalyst prepared according to the method of Section ‘Preparation of catalyst slurry and SCR catalyst’. A fixed-bed reactor was used, and the reactant gas mixture consisted of NO 300 ppm, NH3 300 ppm, O2 5 vol.%, and N2 balance gas with a space velocity (SV) of 60,000 h−1. The outlet gas was monitored by NO–NO2–NOx analyzer (42i–HL, Thermo scientific) and the NOx conversion rate evaluated in the (250 to 450) °C range, the de-NOx efficiency calculated by the following equation:
Results and discussion
Characterization of dispersant and TiO2 slurry
Table 3 summarizes the zeta potential values and pH of each dispersant. The absolute value of the zeta potential was remarkably high at 368WB, followed by BYK110, BYK180, and 369SB. The zeta potential is a value that represents the magnitude of the repulsive force and attractive force between particles charged with positive ( + ) or negative ( − ) charges. Colloids with high zeta potential are electrically stable, while colloids with low zeta potential tend to coagulate or aggregate. 17 Therefore, we confirmed that the 368WB dispersant was the most stably distributed. The pH of BYK110 and 369SB were acidic with values of 1.0 or less and 2.1, respectively. On the other hand, the pH of BYK180 and 368WB were neutral with values of 8.3 and 7.8, respectively. Through the Materials Safety Data Sheet (MSDS) of the products, it was found that BYK110, BYK180, and 368WB are high molecular weight dispersants, while 369SB is a low molecular weight dispersant.
The measurement result of Zeta potential and molecular weight for each dispersant.
Figure 1 reveals the TG and differential TG (DTG) curves of the dispersants. All dispersants, except 369SB, had gradual weight loss from low temperature to the main decomposition regions. Compared to the other dispersants, BYK180 revealed the largest amount of weight loss (about 35%) prior to the main decomposition. The maximum decomposition temperature (Tmax) on the DTG curves of 368WB (376.9°C) was the highest, followed by 369SB, BYK180, and BYK110 at (346.9, 308.1, and 300.3) °C, respectively. This suggests that among the dispersants, 369SB has the highest thermal stability.

Tg and DTG curves of the BYK110, BYK180, 368WB, and 369SB dispersant.
Figure 2 shows the chromatograms of the pyrolysis products generated from the non-isothermal pyrolysis of each dispersant at the main decomposition region. Although peak identification was difficult, all dispersants had acetaldehyde and 1,4-dioxane as the main pyrolyzates. This suggests that the ethylene oxide constitutes the main backbone of all the dispersants. In the case of BYK110 and BYK180, alkanolactones, such as pentanolactone and hexanolactone, were monitored as their main pyrolyzates, suggesting the presence of alkanoates, such as pentanoate and hexanoate, in their polymer structures. Meanwhile, 368WB and 369SB having higher thermal stabilities than BYK110 and BYK180, they also had aromatic hydrocarbons, such as benzene, ethylbenzene, and styrene. This suggests that the higher thermal stability of 368WB and 369SB can be explained by the presence of aromatic groups in their polymer structures. Arnold et al. also explained that the polymers that consisted of aromatic ring had superior thermal and oxidation stabilities. 18 Therefore, it is judged that the Tmax values of 368WB and 369SB were measured relatively higher than that of the other two.

TICs obtained from the Py-GC/MS analysis of dispersants at (260 to 600) °C.
Figure 3 shows the rate of change, calculated as the height of the suspension that changes with time compared to the initial height of the TiO2 slurry described in the Section ‘Characterization of the dispersant and TiO2 slurry’. The sedimentation rate used in the graph was calculated as follows:

Changes in the sedimentation rate of TiO2 slurry according to the type of added dispersant.
According to the sedimentation study, the highest sedimentation rate was shown by TiO2 slurry added with 369SB, and the lowest sedimentation rate was BYK110, followed by BYK180 and 368WB. Previous studies, through the sedimentation study, explained that the lower the change in sedimentation height compared to the sedimentation time, the more stable the dispersion form.19,20 Based on this, it is judged that the nanoparticles in the solvent are most stably dispersed when 368WB is added into the TiO2 slurry. Wei et al. studied the dispersion stability of TiO2 for each type of dispersant in water-soluble isopanol and reported that TiO2 is affected by the pH and zeta potential of the dispersant. 21 It was reported that the dispersibility of TiO2 was optimal when the zeta potential was high and pH ≒ 821, and we obtained similar results in this study.
To evaluate the dispersion shape of the TiO2 slurry according to the type of added dispersant, TEM was used to investigate the dispersibility and aggregation degrees of the nanoparticles in the TiO2 slurry, as shown in Figure 4. The result confirmed that in the case of the slurry to which BYK180 and 368WB were added, the nanoparticles did not aggregate, and were relatively evenly dispersed, while those containing 369SB and BYK110 did not evenly disperse the nanoparticles, and the aggregation phenomenon in the solvent was severe.

Nanoparticles shape in the TiO2 slurry to which BYK110, BYK180, 368WB, and 369SB were added.
According to previous studies, the molecular weight of the dispersant affects the dispersibility of the slurry.22,23 In this study, it was confirmed through Py-GC/MS analysis that the substances constituting the dispersant were very different, and it is judged that the molecular weight of the dispersant composed of a component having a high molecular weight is relatively high. Ohenoja et al. reported that when sodium poly acrylate dispersant with a molecular weight of 12,500 was added to the TiO2 slurry, the viscosity of the slurry was lowered, and the particle size distribution was narrowed. 22 In addition, another study reported that a polymer dispersant with a molecular weight of 30,000 or more significantly decreases the adhesive force between TiO2 particles, and the heavier the molecular weight of the dispersant, the greater the steric repulsive force and the lower the viscosity. 23 Therefore, according to the molecular weight type in MSDS shown in Table 1 and the Py-GC/MS results of the dispersants, the dispersant BYK180 and 368WB can have relatively higher molecular weights than other dispersants, and it is judged that its characteristics help to improve the dispersibility of the slurry.
Characterization of the catalyst slurry and SCR catalyst
Figure 5 shows the results of measuring the viscosity vs. shear rate of the catalyst slurry for each type of dispersant. The catalyst slurry to which 368WB was added has the lowest viscosity characteristics. In the case of catalyst slurry with the remaining three dispersants added, there was no clear difference in viscosity characteristics compared to shear rate. Zhai et al. studied the viscosity characteristics vs. shear rate to confirm the dispersibility of nanoparticles, and used the low viscosity vs. shear rate as an indicator of high dispersibility. 24 Therefore, in this study, it is judged that the slurry containing 368WB, with the lowest viscosity characteristics, exhibited the highest dispersibility.

Rheological behavior of TiO2 catalyst slurry according to the type of dispersant.
As a result of measuring the pH of the catalyst slurry, BYK110, BYK180, 368WB, and 369SB showed values of (3.2, 4.6, 4.1, and 2.6), respectively. The pH of the dispersant itself is shown in Table 3. When the pH of the dispersant itself was strongly acidic, the pH of the catalyst slurry increased slightly, while when the pH of the dispersant was neutral and basic, the pH of the catalyst slurry decreased. Ciambelli et al. investigated the de-NOx efficiency by adjusting the pH of the catalyst slurry with HNO3 and NH4OH, and as a result, confirmed that as the pH of the catalyst increased, the denitrification performance improved. 25 Therefore, if a dispersant having a relatively high pH is used, the pH of the catalyst itself is increased, and this improves the de-NOx efficiency.
Table 4 summarizes the measurement results for the specific surface area, pore size, and pore volume according to the type of dispersant added. Cha et al. reported that the specific surface area was up to 37.6%, the pore size was up to 14.9%, and the pore volume was up to 36.8%, depending on the type of colloidal silica, confirming that colloidal silica affects the modification of the catalyst surface and pore properties. 26 On the other hand, we determined that there is no significant difference in the measurement results of the specific surface area, pore size, and pore volume according to the type of dispersant added.
Specific surface area and porosity of the SCR catalyst.
SCR catalyst de-NOx efficiency evaluation
Figure 6 shows the results of evaluating the de-NOx efficiency of the SCR catalyst by reaction temperature. The catalysts to which each dispersant was added were determined to have high de-NOx efficiency in the range 350°C and 400°C. This is because the V2O5/TiO2 catalysts differ somewhat depending on the additional composition, but most have an active temperature window of (300 to 400) °C, and have high de-NOx efficiency of about 90% or more.27,28 In particular, the overall de-NOx efficiency of the catalyst to which the dispersant 368WB was added, showed high de-NOx efficiency compared with the other dispersants. The catalyst with 368WB showed a slight decrease in performance in the temperature region of 450°C, where normally the catalyst activity decreased. However, in the high-temperature region of 350°C, the catalyst with 368WB showed high activity of 92% or more. In addition, it showed the de-NOx efficiency of over 84% even in the low temperature of 250°C.

The de-NOx efficiency of V2O5/TiO2 catalyst according to temperature variation.
Through the results of Sections ‘Characterization of dispersant and TiO2 slurry’ & ‘Characterization of the catalyst slurry and SCR catalyst’, it was found that compared to the other dispersants, 368WB has relatively high zeta potential, molecular weight, and pH value, and TiO2 slurry with 368WB showed lower viscosity characteristics and lower sedimentation rate. In addition, the de-NOx efficiency was the highest in the catalyst prepared with the TiO2 slurry with high dispersibility, Yan et al. also reported that SCR activity increased when using highly dispersed CuO nanoparticles. 29 It has been reported that the dispersion degree of the catalyst not only increases the activity of the catalyst but also affects the stability.14,30,31 Zhang et al. and Zheng et al. explained that the stability and activity of the catalyst is derived from the particle size and dispersion of the active species.14,30 Howeizi et al. reported that high reducibility of Pd-doped aluminum-based catalysts improved the activity and stability of catalyst, because the reducibility was improved through high dispersion of the catalyst reducing material having a nanoparticle size. 31 The dispersant enhances the dispersibility of the catalyst material, and the great dispersibility of catalyst slurry is considered to have contributed to the improvement of catalyst activity and stability.
Conclusions
In this study, we investigated the characteristics of the dispersant effect on the de-NOx efficiency. The Py-GC/MS results confirmed that the components of the dispersant varied based on the type, and the characteristics of dispersant, such as the zeta potential, pH, and molecular weight, which affected the dispersibility of the catalyst slurry. When a dispersant having a high value of zeta potential, molecular weight, and pH was added into the TiO2 slurry, the dispersibility of slurry was improved, showing a low sedimentation rate and viscosity characteristics. We compared the dispersion shape of TiO2 particles through TEM analysis, and determined that the distribution of particles was even when the dispersibility was excellent. In addition, we measured the activity of the SCR catalyst prepared by adding 1.0 wt.% of different dispersants, and determined that dispersant with excellent dispersibility improved the de-NOx efficiency in the overall temperature range. That is, the zeta potential, pH, and molecular weight of the dispersant affect the dispersibility of the slurry containing TiO2 nanoparticles; also, the dispersibility of the catalyst is judged to have a direct effect on the de-NOx efficiency of the SCR catalyst.
Footnotes
Acknowledgements
This work was supported by the Korea Ministry of Trade, Industry and Energy (MOTIE) as “Development of NOx Removal Catalyst with Wide Temperature Window (No. 20005721).”
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Korea Ministry of Trade, Industry and Energy, (grant number 20005721).
